© Frontier Economics Pty. Ltd., Australia.
Options for scarcity pricing A FINAL REPORT PREPARED FOR SYDNEY CATCHMENT
AUTHORITY
September 2011
i Frontier Economics | September 2011
Contents 11-09-09 Final SCA Scarcity pricing STC Final
Plus Misquotes fixed
Options for scarcity pricing
Executive summary iii
1 Introduction 1
1.1 Study background 1
1.2 Purpose and scope of this paper 1
1.3 Pricing options 2
1.4 Paper structure 2
2 Current water supply and regulatory arrangements 3
2.1 Current metropolitan water supply sources 3
2.2 Variability of dam inflows and levels 5
2.3 Current water security framework 6
2.4 Price regulation 9
2.5 Current water prices and cost structures 11
3 Objectives 13
3.1 What is scarcity pricing trying to achieve? 13
3.2 Principles of efficient pricing of water 14
3.3 Other pricing objectives 18
4 Assessment of options 23
4.1 Overview 23
4.2 Scarcity pricing based on SCA’s operating costs 23
4.3 Scarcity pricing based on the cost of alternative triggered supply and
demand options 29
4.4 Scarcity pricing based on a system optimisation model 35
5 Conclusions 39
References 41
ii Frontier Economics | September 2011
Tables and figures
Options for scarcity pricing
Figures
Figure 1: Sydney water supply system 4
Figure 2: Current and future water supply sources for Sydney 5
Figure 3: Sydney storage levels (%), 1998-2011 6
Figure 4: Change in SCA storages (ML), 1998-2011 6
Figure 5: Stylised representation of operating rules based on dam levels 9
Figure 6: IPART Determination process 10
Figure 7: Current prices and short-run costs related to dam trigger levels 12
Figure 8: Social SRMC (including the opportunity cost of water) 16
Figure 9: Demand management and water savings in Sydney, 2000 to 2010.
19
Figure 10: Prices under scarcity-based pricing, across various rainfall
scenarios 36
Tables
Table 1: Illustrative scarcity pricing schedule to Sydney Water (based on SCA
operating costs) 24
Table 2: Indicative revenues from volumetric and fixed charges under
different pricing regimes and dam levels 26
Table 3: Illustrative scarcity pricing schedule to Sydney Water - cost of
alternative triggered supply and demand options 30
Table 4: Illustrative scarcity pricing schedule to Sydney Water (smoothed
prices) 30
Table 5: Indicative revenues from volumetric charges under different pricing
regimes and dam levels 32
September 2011 | Frontier Economics iii
Executive summary
Executive summary
Background
Scarcity pricing has figured prominently in the public debate on water pricing,
particularly since the widespread imposition of severe water restrictions across
urban Australia in recent years.
The basic idea of scarcity pricing is to set volumetric prices to reflect the
opportunity costs of using water storages as dam levels change. For example, the
price would be relatively low when the dam is full (and the probability of running
out of water is low) and higher when storages decline (and there is a need to
augment dam water with emergency supplies or impose restrictions).
In its 2009 price determination for Sydney Catchment Authority (SCA), the
Independent Pricing and Regulatory Tribunal (IPART) expressed interest in
possibly developing a form of ‘scarcity pricing’ for potential implementation at
the 2012 SCA price determination, particularly for wholesale prices (i.e. SCA
charges to Sydney Water).
In the context of pricing at the wholesale level, one of the key objectives is
encouraging efficient water use and investment by Sydney Water and facilitating
competition in bulk supply.
This paper examines a small number of potential options and assesses them
against key pricing objectives taking into account SCA’s strategic interests, as well
as the broader implications for the metropolitan system. We compare each
option to the status quo situation (of set fixed and variable bulk water charges) in
both wet and dry inflow sequences.
Our analysis
Under current pricing arrangements there is an inherent conflict between
achieving revenue stability for SCA on the one hand and IPART’s desire to send
a signal on the value of water to inform Sydney Water’s sourcing and investment
decisions on the other. A key issue in assessing options for scarcity pricing is
whether they can provide appropriate signals for efficient use and investment
while not exposing SCA to undue revenue risk. There is a strong argument that
SCA, as a relatively passive manager of catchments and dams in accordance with
government defined operating strategies, should be able to recover its efficient
costs without being excessively exposed to demand risk over which it has no
control.
Our view is that scarcity pricing (based on principles of marginal cost pricing)
would address this issue. This model would broadly involve:
● Aligning SCA’s volumetric price with its short-run operating costs, with a
fixed charge to address any revenue shortfall. This would require a large
iv Frontier Economics | September 2011
Executive summary
increase in revenue generated from SCA’s fixed charge compared to current
arrangements.
● Setting a separate volumetric price that reflects the estimated marginal value
of water in storage, which would be in addition to SCA’s infrastructure
charges but would effectively apply only when predefined triggers are
reached.
The revenue collected through this additional scarcity charge represents a
separate resource rent outside of SCA’s required revenue requirement
and could potentially be retained by government or alternatively used to
offset Sydney Water’s fixed charges in the current or future regulatory
periods.
This pricing approach would reduce the revenue risk to SCA embodied in
current tariff structures while providing an appropriate price signal for
consumption and investment to Sydney Water (and other SCA customers) and
potential new entrants at times when water is scarce. It does so by clearly
differentiating between pricing for SCA’s infrastructure services and the water
resource itself.
The extent to which these prices will have a material effect on the current and
future portfolio of supply options for Sydney will partly depend on broader
institutional arrangements for urban water planning. For example, current
government policies (for example, strict desalination plant operating rules) may
lock in particular sourcing decisions and investments and thus limit Sydney
Water’s flexibility to respond to wholesale water prices.
The design of the scarcity pricing regime also needs to consider carefully how
scarcity prices will combine with existing operating rules for the Sydney system to
achieve the most efficient mix of options for balancing supply and demand. For
example, to the extent the operating rules for the desalination plant already take
into account risks to existing storages and therefore reflect an ‘optimal’ operating
strategy, there is a risk that the addition of scarcity pricing for SCA supplies will
double count the costs of consuming dam water. This is not so say scarcity
pricing is not worthwhile, but rather that current operating rules may need to be
reconsidered in light of this new option for balancing supply and demand.
Putting this pricing model into practice will be challenging. In particular,
estimating the value of water in storage is a key issue to determine. In this paper
we canvass potential options for estimating the marginal value of water in storage
ranging from heuristic approaches based on existing operating rules (e.g. setting
prices equal to the operating cost of alternative options such as desalination when
dam levels trigger operation of the desalination plant) to economic modelling
approaches.
In theory, an economic model that calculated an optimal price based on existing
system constraints, planned investments and operating plans would produce
September 2011 | Frontier Economics v
Executive summary
efficient price signals. However, we recognise that much work would be required
to develop such a model and for it to be accepted in a regulatory price setting
context.
Conclusions
In summary, we would advocate replacing the current pricing arrangements
based on setting SCA’s volumetric charges with regard to its LRMC with a more
cost-reflective approach based on SCA’s SRMC together with an additional
scarcity price based on the costs of predefined triggered alternatives. This would
better protect SCA’s revenue adequacy while also achieving IPART’s aim of a
more efficient price signal to SCA’s customers and potential new suppliers.
September 2011 | Frontier Economics 1
Introduction
1 Introduction
1.1 Study background
Scarcity pricing has figured prominently in the public debate on water pricing,
particularly since the widespread imposition of severe water restrictions across
urban Australia in recent years.
While a number of variants have been proposed, the basic idea of scarcity pricing
is that the price of water would be higher when water was relatively scarce (e.g.
dam levels were low) and lower when water was more plentiful (e.g. when dam
levels were high). The underlying rationale for scarcity pricing is that it may be a
more efficient way of balancing supply and demand, particularly for short-term
shortages, and could signal the cost of using rainfall-dependent sources of supply.
Scarcity pricing could potentially occur at the wholesale level and/or retail level.
In its 2009 Determination for the SCA, IPART decided not to implement retail
scarcity pricing at that time but canvassed the idea of introducing a form of
scarcity pricing at the wholesale level.
It flagged that it was interested in receiving stakeholders’ views on the potential
application of scarcity pricing in Sydney and in particular on the design and
application of such a pricing model, implementation issues to be addressed, and
its potential advantages and disadvantages.
Scarcity pricing at the wholesale level would obviously have major implications
for the SCA, particularly in terms of its recovery of costs and revenue volatility in
the context of a regulated price path. A wholesale scarcity price would also have
broader implications for other stakeholders, agencies and customers in relation to
the efficient optimisation of the portfolio of supply and demand side measures
contributing to supply security and reliability in the Sydney metropolitan area.
1.2 Purpose and scope of this paper
Against this background, the key deliverable from the consultancy is a concise
paper that identifies and assesses a range of wholesale scarcity pricing options for
strategic discussion by the Board, prior to discussion with IPART.
The paper examines a small number of potential options and assesses them
against key pricing objectives taking into account SCA’s strategic interests, as well
as the broader implications for the metropolitan system. We compare each
option to the status quo situation (of set fixed and variable bulk water charges) in
both wet and dry inflow sequences.
The project was largely a desktop exercise with consultation with the SCA’s
project manager as required.
2 Frontier Economics | September 2011
Introduction
1.3 Pricing options
We assessed the following broad approaches to setting SCA’s wholesale water
charges:
● Current approach or status quo - SCA recovers costs from Sydney Water
using a two-part tariff (i.e. fixed and variable charge), with IPART setting the
variable charge with reference to the long-run marginal cost (LRMC) of
supply. There is no scope to adjust variable charges during the regulatory
period in response to sudden reductions in dam levels or associated increases
in supply costs.
● Scarcity pricing based on SCA’s operating costs – this involves setting
SCA’s variable charge to Sydney Water based on its short-run operating costs
and increasing this charge when dam levels trigger increased operating costs
(particularly due to Shoalhaven pumping cost).
● Scarcity pricing based on the cost of alternative triggered supply and
demand options – this involves setting a variable charge to Sydney Water
based on estimates of the opportunity cost of using dam water (e.g. the cost
of Sydney Water operating the desalination plant or imposing water
restrictions). The price increases would have links to existing operating rules
that require Sydney Water to deploy supply or demand management options
when dam levels fall to a certain levels.
● Dynamically efficient pricing based on a system optimisation model
for Sydney – involves using an economic model to calculate a schedule of
efficient prices defined in terms of dam levels.
Consistent with IPART’s proposal, we examined these pricing options on the
basis that they would apply in conjunction with existing institutional and policy
settings (e.g. desalination operating rules, restrictions policies etc) rather than as
an alternative. Our analysis also assumes that scarcity pricing applies only at the
wholesale level and not the retail level. However, we do identify the implications
of these constraints for the efficacy of scarcity pricing options.
1.4 Paper structure
The structure of the rest of this paper is as follows:
● Section 2 describes current water supply and regulatory arrangements.
● Section 3 describes pricing objectives.
● Section 4 assesses the pricing options compared to the status quo.
● Section 5 provides our conclusions.
September 2011 | Frontier Economics 3
Current water supply and regulatory
arrangements
2 Current water supply and regulatory
arrangements
This section provides background information on current water supply and
regulatory arrangements in Sydney, which is relevant to the design and
implementation of scarcity pricing options at the wholesale level.
2.1 Current metropolitan water supply sources
SCA and Sydney Water are the two main water businesses responsible for
operating Sydney’s water supplies. SCA is responsible for catchment operations
and selling bulk water supplies to Sydney Water (as well as three local councils)1
from its system of dams (including transfers from the Tallowa dam on the
Shoalhaven River)(Figure 1). Sydney Water supplies retail water supplies to
metropolitan customers drawing on raw water supplies from SCA and other bulk
water sources (i.e. desalination, recycled water).
1 Shoalhaven City Council, Goulburn Mulwaree Council and Wingecarribee Shire Council.
4 Frontier Economics | September 2011
Current water supply and regulatory
arrangements
Figure 1: Sydney water supply system
Source: Sydney Catchment Authority.
September 2011 | Frontier Economics 5
Current water supply and regulatory
arrangements
Sydney Desalination Plant Pty Limited (SDP), a wholly owned subsidiary of
Sydney Water, owns the Sydney desalination plant. The private sector operates
and maintains the desalination plant facilities in return for monthly performance-
based payments based on a formula that includes variable costs associated with
the daily drinking water volumes from the plant (Sydney Water, 2011).
Although the NSW government has taken steps to diversify Sydney’s water
supplies in the past decade by investing in desalination and recycling, the majority
of Sydney’s water supply still comes from capturing rainwater and storing it in
dams (NSW Office of Water, 2010). In 2011, water available from dams was 570
GL per year compared to 90 GL from desalination (Figure 2). The Metropolitan
Water Plan for Sydney (NSW Office of Water, 2010) indicates that increases in
recycling, desalination plant capacity and improvements in water-use efficiency
could help meet future demand needs.
Figure 2: Current and future water supply sources for Sydney
Source: NSW Office of Water, 2010.
2.2 Variability of dam inflows and levels
Although the capacity of Sydney storages is one of the largest in the world per
head of population (NSW Office of Water, 2010) inflows are highly variable and
dam levels can increase or decrease significantly from year to year. Since the early
1990s, average rainfall and inflows have declined compared to previous decades
(1950s to early 1990s) and there have been substantial drought transfers from the
Shoalhaven (CIE 2010). Examples of the variability of dam levels in Sydney
include the decline in storages from 90% to approximately 30% between 2001
and 2007 (Figure 3). In the eight months leading up to March 2009, dam levels
fell from above 65 percent to 58 percent (SCA, 2009).
6 Frontier Economics | September 2011
Current water supply and regulatory
arrangements
These fluctuations in dam levels are greater than in other major cities. Although
Melbourne also experienced large reductions in inflows between 2001 and 2007,
for example, dam levels decreased from around 60% to 40% during that time
(Melbourne Water, 2010).
Figure 3: Sydney storage levels (%), 1998-2011
Source: SCA 2011
Figure 4 presents storage levels in volumetric terms. It shows that dam level can
decrease several hundred GL (1GL = 1000ML) in one year.
Figure 4: Change in SCA storages (ML), 1998-2011
Source: SCA 2011
2.3 Current water security framework
2.3.1 Responsibilities for supply planning and operations
New South Wales has adopted a standing committee approach to formulate a
water security framework for metropolitan Sydney. Chief executive officers from
all water businesses and key government departments make up this committee,
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Jan-9
8
Jul-
98
Jan-9
9
Jul-
99
Jan-0
0
Jul-
00
Jan-0
1
Jul-
01
Jan-0
2
Jul-
02
Jan-0
3
Jul-
03
Jan-0
4
Jul-
04
Jan-0
5
Jul-
05
Jan-0
6
Jul-
06
Jan-0
7
Jul-
07
Jan-0
8
Jul-
08
Jan-0
9
Jul-
09
Jan-1
0
Jul-
10
Jan-1
1
Sto
rage (
%)
Available storage
Current available storage = 74.5%
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
1,600,000
1,800,000
2,000,000
2,200,000
2,400,000
2,600,000
2,800,000
Jan-9
8
Jul-
98
Jan-9
9
Jul-
99
Jan-0
0
Jul-
00
Jan-0
1
Jul-
01
Jan-0
2
Jul-
02
Jan-0
3
Jul-
03
Jan-0
4
Jul-
04
Jan-0
5
Jul-
05
Jan-0
6
Jul-
06
Jan-0
7
Jul-
07
Jan-0
8
Jul-
08
Jan-0
9
Jul-
09
Jan-1
0
Jul-
10
Jan-1
1
Megalitr
es
Total system full operating storage
Available storage
50% Operating storage
Current available storage = 74.5%
September 2011 | Frontier Economics 7
Current water supply and regulatory
arrangements
with the NSW Office of Water acting as a secretariat. The Minister for Water is
responsible for approving the Metropolitan Water Plan. An independent panel,
comprising specialists in environmental management, economics, social research
and water industry experts oversee the planning process (PwC, 2010). The NSW
government maintains ultimate control over the mix of measures to secure
greater Sydney’s water supply.
Under the metropolitan plan, Sydney Water is responsible for a wide range of
initiatives, including wastewater recycling, desalination, demand management and
leak reduction. Sydney Water’s operating licence specifies water efficiency targets,
demand management and recycling requirements issued by the Government.
Sydney Desalination Plant (a subsidiary of Sydney Water) holds a Retail Supplier
and a Network Operator licence under the Water Industry Competition Act
2006, which include operating rules for the desalination plant issued by
Government (Sydney Water, 2011).
SCA holds a water management licence under the Water Act 1912, which
specifies rules for operating the Sydney bulk supply system, including
environmental flows and Shoalhaven transfers. Its operating licence includes
provisions relating to infrastructure management and water conservation
including undertaking practicable actions to conserve water and minimise water
losses, which may include working collaboratively with its customers (SCA,
2011).
2.3.2 Supply operating rules
Supply operating rules govern how the SCA and Sydney Water operate the
metropolitan supply system. Specific rules in the metropolitan plan include:
● Shoalhaven transfer rules (for SCA): under system operating rules,
transfers from Tallowa Dam in the Shoalhaven can begin when Sydney’s total
dam storage level falls below 75 percent but only while the storage level of
Tallowa Dam is above its minimum operating level of minus one metre from
full supply level.2 In severe drought, the plan allows the minimum operating
level for transferring water from Tallowa Dam to Sydney to lower to minus
three metres (NSW Office of Water, 2010).3 The SCA must cease water
transfers from the Shoalhaven system when total system storage reaches 80%
(SCA Water Management Licence).
2 The full supply level is the level of the water surface in storage when it is at its maximum operating
level under normal conditions (not flood conditions). The minimum operating level helps ensure
‘the Shoalhaven community’s water supply is secure and the health of the lower Shoalhaven River
system is maintained with ongoing environmental flows’ (NSW Office of Water, 2010).
3 SCA’s water management licence states ‘The SCA must not commence transferring water from the
Shoalhaven system via the Shoalhaven Scheme unless total system storage is less than 75%’.
8 Frontier Economics | September 2011
Current water supply and regulatory
arrangements
● Desalination operating rules (for Sydney Water): the desalination plant
will run at full capacity (i.e. 90 GL/year) during a two-year ‘defects correction
period’, which will end in mid June 2012. After this period, the plant will
operate at full production capacity and supply desalinated water to Sydney
Water’s area of operations when the total dam storage level is below 70 per
cent and will continue to do so until the total dam storage level reaches 80
per cent (NSW Office of Water, 2010). The metropolitan plan notes ‘if
necessary, the Government will be able to operate the desalination plant at
other times to secure water supplies (for example if availability of water from
other parts of the supply system were affected by technical or other
problems)’ (NSW Office of Water, 2010). As an input to developing the
Metropolitan Water Plan, Sydney Water commissioned the Centre for
International Economics (CIE) to assess the net benefits of different
operating regimes for the desalination plant.4 This review considered three
alternative operating rules (30/40, 70/80 and 80/90)5 and recommended the
‘70/80 rule’ above based on assumptions about other aspects of system
management, such as restrictions policies (CIE , 2010).
● Drought restrictions (enforced by Sydney Water): In 2010, the NSW
government announced a revised mandatory restrictions regime, made up of
two levels commencing at around 50 percent and 40 percent of Sydney’s total
dam storage levels. Sydney’s total dam storage level, predicted weather
patterns, the season, and demand forecasts will influence the exact timing for
introducing drought restrictions (NSW Office of Water, 2010). Sydney
Water’s operating licence notes it may place conditions on water use by
customers at the discretion of the Minister or Government.
Figure 5 presents a stylised representation of the operating rules for the Sydney
supply system. The triggers for commencing and ceasing operation of particular
supply options and restrictions may differ. Some triggers are binding (e.g. Sydney
Water must run the desalination when storages fall to 70%), while others are
more flexible (e.g. SCA may transfers water when storages fall to 75%).
4 The estimate of net benefits included the costs of operating the desalination plant, the costs of water
restrictions, avoided infrastructure costs, supply security benefits, and environmental impacts.
5 The first number is the storage level at which Sydney Water switches the desalination plant on and
the second number is the storage level at which Sydney Water switches the plant off.
September 2011 | Frontier Economics 9
Current water supply and regulatory
arrangements
Figure 5: Stylised representation of operating rules based on dam levels
Source: Trigger levels based on 2010 Metropolitan Water Plan (NSW Office of Water, 2010)
2.4 Price regulation
SCA and Sydney Water are subject to economic regulation by IPART. Figure 6
presents the broad framework for IPART’s determination process.
Currently, IPART uses a building block approach to calculate SCA’s notional
revenue requirement. To apply this approach, it makes decisions on the revenue
SCA will require for efficient operating expenditure and capital investment over
the determination period (which is currently three years - July 2009 to 30 June
2012). It then considers appropriate price levels and prices structures taking into
account objectives such as protecting SCA’s financial viability, encouraging
economic efficiency and protecting water consumers from price shocks (IPART,
2009).
0%
20%
40%
60%
80%
100%
10%
30%
50%
90%
70%
Shoalhaven transfers /desalination trigger off
Shoalhaven transfers on
Desalination on
Level 1 restrictions on
Level 2 restrictions on
Storage levels
Restrictions trigger off (above level for 12 months)
Time
Storage capacity
10 Frontier Economics | September 2011
Current water supply and regulatory
arrangements
Figure 6: IPART Determination process
Source: IPART 2009
IPART can adjust SCA’s notional revenue requirement to account for
unexpected developments during the previous regulatory period, such as
differences between actual and forecast water sales or capital expenditure. It can
also include regulatory mechanisms to address the risk of variations between
actual and forecast required revenue in the upcoming regulatory period (such as
by allowing SCA to pass through unexpected costs associated with Shoalhaven
transfers). In the most recent determination in 2009, IPART did not allow a cost
pass-through mechanism for Shoalhaven transfers as it deemed such transfers
were unlikely during the regulatory period following the government imposing a
temporary moratorium of such transfers (IPART, 2009).
September 2011 | Frontier Economics 11
Current water supply and regulatory
arrangements
2.5 Current water prices and cost structures
In general, the short-run marginal cost (SRMC) of water supply is the cost of
providing a unit of water when infrastructure capacity is fixed and the long run
marginal cost (LRMC) of supply is the cost of providing a unit of water when it is
possible to vary infrastructure capacity. Regulators sometimes set volumetric
prices for water to reflect one of these costs.
Figure 7 presents current prices and indicative cost structures facing SCA and
Sydney Water and relates these to storage levels6. It shows SCA’s current
volumetric price for raw water (approximately $250/ML) is above SCA’s short-
run marginal cost (SRMC) (up to $70/ML) but below the SCA’s LRMC (at least
$1200/ML)7. As recent estimates of the SCA’s SRMC are based on pumping
costs from the Shoalhaven (i.e. $70/ML), which are additional costs incurred by
SCA when storages are below 75%, we have assumed a hypothetical SRMC of
$30/ML to represent SCA’s SRMC without Shoalhaven transfers.
Sydney Water’s short-run cost when dam levels are above 70% is essentially
SCA’s raw water charge ($250/ML) plus additional filtration cost ($80/ML) (CIE
2010). The marginal operating cost of desalination is at least $422/ML8 (IPART,
2009). IPART has previously estimated Sydney Water’s LRMC as $1.93/kL (or
$1930/ML) based on the cost of expanding the capacity of the existing
desalination plant from 250ML/day to 500ML/day (Independent Advisory
Panel, 2008).
In 2005, IPART increased the proportion of revenue SCA obtained from its
volumetric charges to Sydney Water from half to two-thirds. It did this to ‘send a
pricing signal to Sydney Water that will help achieve the State Government’s
demand management objectives’ and help achieve the objective of setting charges
with reference to SCA’s Long Run Marginal Cost of supply (LRMC)(IPART,
2009).
6 It is recognised that these prices and cost structure may increase in the future due to factors such as
changes in energy prices. The numerical analysis in this report is for illustrative purposes rather than
representing expectations of actual outcomes.
7 In general terms, SCA’s LRMC is calculated as the present value of the cost of SCA’s next supply
augmentation measure divided by the present value of the amount of water supplied by the measure.
IPART based its estimate of SCA’s LRMC ($1200/ML) on indicative estimates of the cost and yield
of SCA’s next likely supply augmentation project (a form of Shoalhaven transfers project) (IPART,
2009).
8 CIE estimated additional water is produced by desalination at approximately 60 cents per kL or
$600/ML (excluding start-up and shutdown costs)(CIE 2010).
12 Frontier Economics | September 2011
Current water supply and regulatory
arrangements
Figure 7: Current prices and short-run costs related to dam trigger levels
Source: Triggers levels based on 2010 Metropolitan Water Plan (NSW Office of Water, 2010)
*Estimated as at least this amount # Hypothetical.
Desal. Cost (Sydney Water)
SCA Bulk price
SCA SRMC (drought transfers)SCA SRMC (no drought transfers #)
7030
250
422
Sydney Water LRMC1930
$/ML
1200 SCA LRMC *
Sydney Water bulk cost (Bulk price + filtration) 330
Dam levels (%)60 50
Sydney Water SR Cost Curve
100 90 80 70
SCA SR Cost Curve
September 2011 | Frontier Economics 13
Objectives
3 Objectives
3.1 What is scarcity pricing trying to achieve?
Currently, SCA’s volumetric price to Sydney Water for raw water does not vary
with short-term changes in water storages; the price is the same regardless of
whether the dam is half full or spilling. A key concern about this approach is that
the volumetric price does not send appropriate signals to Sydney Water about the
true cost of using dam water as storages decline and can result in unintended
consequences. These include triggering water restrictions more frequently (which
imposes cost on water customers) or inhibiting investment in other supply
options by underpricing SCA’s raw water.
The basic idea of scarcity pricing is to set volumetric prices to reflect the
opportunity costs of using water storages as dam levels change. For example, the
price would be relatively low when the dam is full (and the probability of running
out of water or imposing restrictions is low) and higher when storages decline
(and there is a need to augment dam water with emergency supplies or impose
restrictions).
In its 2009 price determination for SCA, IPART expressed interest in possibly
developing a form of ‘scarcity pricing’ for potential implementation at the 2012
SCA price determination, particularly for wholesale prices (i.e. SCA charges to
Sydney Water) (IPART, 2009).
In the context of pricing at the wholesale level, one of the key objectives is
encouraging efficient water use and investment by Sydney Water and facilitating
competition in bulk supply. IPART (2009), for example, have expressed interest
in scarcity pricing on the basis it may:
- provide incentives to Sydney Water to invest in water conservation and demand
management measures, where efficient
- signal to Sydney Water when it is more appropriate to obtain water from sources
other than SCA, and vice-versa
- provide signals to potential new suppliers of bulk water as to when it may or may
not be viable for them to invest in new water supply infrastructure
IPART clearly wishes to explore the scope for scarcity pricing to assist in the
optimal utilisation of and investment in the range of alternative sources of
supply. This appears to reflect an underlying concern that if SCA prices do not
reflect the underlying value of the water itself, there may be an incentive for
Sydney Water (and other SCA customers) to use too much of water from
storages and undermine future security of supply, as opposed to using
desalination or recycled. IPART (2009) note:
14 Frontier Economics | September 2011
Objectives
‘While acknowledging the dominant role that SCA is likely to continue to play in the
provision of water, IPART considers that it is also important to recognise that Sydney
is likely to increasingly have alternative sources of water supply. In addition to SCA’s
dams, desalination and the use of recycled water for non-potable purposes will
become increasingly important. Some alternative sources of water are owned by
Government, but others may be privately owned. In these circumstances, it is worth
investigating the role that pricing can play in providing effective signals to both
Sydney Water and potential new suppliers of bulk water, to ensure that Sydney’s
water needs are supplied at least cost to the community.’
In theory, scarcity pricing at the wholesale level could feed through to the retail
prices of Sydney Water and provide signals to consumers of the resource to
reduce consumption in times of scarcity. However, scarcity pricing at the
wholesale level does not require water charges to Sydney Water customers vary
with dam levels. In fact, IPART (2009) notes:
A separate question is whether Sydney Water’s retail prices should also vary with
SCA’s dam levels to reflect the economic value of water. IPART notes that this does
not necessarily need to occur, even if SCA’s wholesale price to Sydney Water does
vary with dam levels.
If a form of scarcity pricing were to be introduced at the retail level, IPART envisages
that it would be applied at the margin, targeting discretionary water consumption only
and operating to support the water restriction regime in equating water demand with
supply.
As noted earlier, the focus of this paper is scarcity pricing at the wholesale level.
3.2 Principles of efficient pricing of water
The objectives espoused by IPART are essentially ones relating to the concept of
economic efficiency, which requires allocating resources across all consumption
and production activities (present and future) in a manner that maximises
benefits to society.
Theory of marginal cost pricing
Economic theory suggests pricing resources at social marginal cost (defined as
the cost of meeting an incremental increase in demand for water) provides an
efficient basis for allocating resources.
Griffin (2006) argues that the marginal cost of supply potentially can include:
● short-run infrastructure operating costs - these costs typically encompass
pumping and treatment costs that vary with output. Short-run operating costs
can increase during drought as the water business uses higher cost supply
options.
September 2011 | Frontier Economics 15
Objectives
● the marginal value of water - the value of an extra unit water from renewable
natural water sources to society.9 This increases if water availability decreases.
If water is not scarce, however, the marginal value of water will be zero.10
● the marginal capacity costs – the social costs incurred when infrastructure
capacity (e.g. storage and delivery infrastructure) is fully employed and the
quantity of capacity demanded exceeds the quantity supplied. For example,
over time population growth may mean that a city’s annual water requirement
increases relative to the supplier’s delivery capacity. At the extreme, the
supplier may need to ration supply capacity thus imposing costs on
customers. The marginal capacity cost will only be non-zero when
infrastructure capacity is constraining (e.g. capacity constraints in delivering
water). In practice, investments in infrastructure capacity are typically lumpy
and generally avoid infrastructure constraints that require rationing of
capacity. Hence, the marginal value of capacity in this context is defined as
the incremental cost to the water business of installing additional
infrastructure capacity to meet supply needs (i.e. the basis of LRMC pricing).
The estimates of short-run marginal cost in section 2.5 only relate to
infrastructure operating cost incurred by SCA in undertaking its supply activities.
However, from society’s perspective the marginal cost of water supply may
exceed operating costs as either water becomes scarce (the marginal value of
water increases) or infrastructure capacity becomes scarce and limits supply (the
marginal capacity cost increases). The marginal value of water (or marginal user
cost) rises as water availability/dam levels fall.
Figure 8 provides an example of how the social marginal cost of supply might
vary with dam levels. As dam levels fall, SCA incurs additional operating costs
associated with managing drought supplies (i.e. Shoalhaven transfers). At the
same time, the opportunity cost of using dam storages (marginal value of water)
gradually increases. Hence, the overall increase in the social marginal cost of
water as dam levels fall reflects both an increase in operating costs of delivering
the water and an increase in the value of water itself. We assume that there are no
infrastructure constraints and the marginal capacity cost for SCA is zero. Based
on this broader definition of the SRMC, SRMC pricing could exceed the current
volumetric bulk charges.
In practice, the marginal value (opportunity cost) of water is notoriously difficult
to estimate as it is not readily observed in a market, as in the rural sector. As
9 Where a source is depletable, Griffin (2006) suggests the appropriate measure is the marginal user
cost – the present value of an extra unit of water in the future. Griffin suggests this mostly relates to
groundwater but could apply to a dams containing tight supply over the next several periods.
10 Griffin assumes this water has a zero accounting costs to the water business (i.e. it is not purchased
in a market).
16 Frontier Economics | September 2011
Objectives
discussed below, stakeholders have proposed a variety of approaches to estimate
the marginal value of water ranging from rules of thumb to sophisticated models.
Figure 8: Social SRMC (including the opportunity cost of water)
Despite the efficiency properties of short run marginal cost pricing, regulators
have typically not applied SRMC but rather have tended to have regard to
LRMC. For example, IPART (2004) has previously noted:
The Tribunal considers that the appropriate next step towards wholesale water price
reform is to review the balance between the fixed access charge and the variable
usage charge and, if possible, set the usage charge with reference to the SCA’s long
run marginal cost. Long run marginal cost here refers to the additional costs of the
measures that the SCA must incur to balance supply and demand, divided by the
amount of additional water provided by those measures.
The LRMC approach to pricing has tended to focus on providing a smoothed
long-term pricing signal to customers. An implicit assumption is that the service
is being provided by a monopoly supplier where its availability is determined
solely by the supply capacity which will need to be augmented when demand
grows to take up all of the existing capacity. While this may be a reasonable
assumption for many services – and for water when the supply is reliable, it is
increasingly recognised that this may not hold for water supply in Australia.
Pricing at long-run marginal cost communicates the expected cost of consuming
an additional unit of water. If the expectations underpinning the calculation of
long-run marginal cost are accurate, then the incentives created for water use and
conservation will be efficient. However, given uncertainty about future demand,
SCA operating costs70
30
250
$/ML
Dam levels (%)60 50100 90 80 70
Opportunity cost of using water in storage (not priced)
Social SRMC (operating cost + opportunity cost of water)
Current bulk charge
September 2011 | Frontier Economics 17
Objectives
the efficient investment path, and particularly supply from rainfall-dependent
water sources, these expectations will almost certainly not be accurate. Periods of
short-run scarcity are likely, and during these times long-run marginal cost pricing
will fail to represent accurately the increases in the opportunity cost of water
stemming from that scarcity.
Further, given that Sydney Water has alternative options for meeting demand
other than SCA dam water there is a question of whether SCA’s LRMC is
appropriate or relevant for informing Sydney Water’s water sourcing and
investment decisions or as a basis for efficient competition from new suppliers.
In particular:
● SCA’s LRMC of infrastructure may be problematic for sending efficient
signal for consumption, particularly given SCA only has some of the available
supply augmentation options. That is, Sydney Water may well have
alternative options such as recycling that it would take up prior to any
augmentation of SCA’s system.
● If Sydney Water bases its decisions about operating the desalination plant or
deploying demand management options according to their short-run
marginal cost, pricing SCA’s supplies to Sydney Water at LRMC will distort
Sydney Water’s sourcing decisions (i.e. it may produce too much desalinated
water compared to drawing on dam water).
● Similarly, in a competitive market, prices will tend to short run marginal cost.
While pricing SCA water at LRMC may make potential alternative supply
sources more attractive, this would not represent efficient competition.
● Setting prices with respect to LRMC requires a great deal of information to
estimate future demand and supply conditions, in order to plot the efficient
path for investment. Consequently, there may be a great deal of uncertainty
about whether such prices reflect the efficient price.
The Independent Panel (2008) notes that changes in the water supply
environment may require rethinking the current approach of LRMC pricing:
In light of quite dramatic and ongoing change to the nature of water supply
businesses in Sydney, there is a risk that price determinations made along existing
guidelines might fail to keep pace with change
The impact of drought and climate change is indeterminate and still unfolding. It has
influenced the introduction of readiness options, one important one of which has
been exercised in the form of the desalination plant.
The existence of this source raises questions about using LRMC as a basis for bulk
water pricing in particular.
The advent of the Water Industry Competition Act means that there are potentially
many smaller water suppliers to enter the market in Sydney, which changes the
landscape in comparison to the earlier model of a single monopoly provider of both
bulk and retail water.
18 Frontier Economics | September 2011
Objectives
Sending suitable price signals to consumers, and maintaining consistency of signals
through the supply chain becomes more of a challenge in this environment.
In practice (as seen in the previous chapter), SCA’s volumetric prices is between
the two extremes of the SCA’s short-run marginal operating costs (i.e. the SRMC
when water and infrastructure capacity are not scarce) and an estimate of SCA’s
long-run marginal cost. This perhaps reflects attempts to compensate for the lack
of an explicit price signal for the marginal value of water.
Benefits of efficient pricing
Scarcity pricing could provide a more cost-reflective signal to Sydney Water
regarding the cost of drawing on dam water during drought (i.e. increased
operating costs and the marginal value of water). As dam levels fall, for example,
the increase in price for dam supplies would provide Sydney Water with an
incentive to substitute dam water with other sources (e.g. recycling, demand
management) which would in turn take pressure off dam supplies and reduce the
risk of restrictions.
Hence, guiding questions relevant to assessing specific pricing options include:
● Do pricing arrangements provide Sydney Water with incentives to source
supplies and invest in demand management in manner that maximises net
social benefit?
● Does the approach provide incentives for efficient competition in the
provision of water services?
3.3 Other pricing objectives
There is a number of other pricing objectives, including:
● Effectiveness
● Revenue adequacy and stability
● Appropriate allocation of risk
● Administrative simplicity
● Transparency.
Effectiveness
While scarcity pricing may provide more efficient sourcing and investment
signals to Sydney Water, its effectiveness in influencing Sydney Water’s sourcing
and investment decisions is another matter. SCA (2009) notes IPART has
previously argued that financial incentives are not as effective for a Government
owned business such as Sydney Water which respond to other imperatives.
September 2011 | Frontier Economics 19
Objectives
Conceptualising Sydney Water’s operating decisions as being strongly influenced
by current wholesale prices can be overly simplistic for several reasons. As noted
above, for example, Sydney Water has licence obligations that require it to meet
specific supply standards and/or invest in recycling and demand management.
These obligations internalise, or at least reduce, the risks associated with water
restrictions even in the absence of price signals reflecting the marginal value of
water. 11 Hence, although the marginal cost of recycling and demand management
projects can significantly exceed the current raw water price, investment in
recycling and demand management in Sydney has nevertheless grown over the
past decade (Figure 9).
Figure 9: Demand management and water savings in Sydney, 2000 to 2010.
Source: Sydney Water, 2010.
Similarly, the operating rules for the Sydney Water desalination plant (developed
by government) currently limit Sydney Water’s flexibility to change its operations
in to response to price signals from SCA. Hence, the likely effect of scarcity
pricing on Sydney Water’s bulk water sourcing decisions needs to be considered
within the broader institutional setting
11 Modelling underlying the metropolitan plan presumably places a social value on different options
based on their contribution to supply security.
20 Frontier Economics | September 2011
Objectives
A further point to note is that Sydney Water knows that it will incur higher costs
as dam levels fall and thus it will have a financial incentive to undertake actions to
mitigate this risk even in the absence of an explicit price on dam water.
Guiding questions relevant to assessing specific pricing options include:
● Does Sydney Water have the capacity to respond to the price signal from
SCA?
● Is the level of the price signal likely to be material?
Revenue adequacy and stability
Pricing principles adopted at the state and national levels recognise the principle
that regulated business should be able to recover efficient costs of supplying
services (Council of Australian Governments, 2004 and 2010).
In a regulated setting, revenue adequacy relates to ensuring all legitimate costs are
included in the revenue requirement and the pricing structure provides the
business with reasonable opportunity to recover this revenue.
Typically, regulated businesses recover required revenue through a two-part
tariff, with the volumetric charge designed to send an appropriate signal to
customers about the ‘marginal costs’ of additional consumption. Because pricing
at marginal cost (particularly where this is low) may not generate sufficient
revenue to recover total costs, the fixed charge is designed to recover the balance
of the revenue requirement (based on estimates of likely demand and hence
revenue likely to be generated from the volumetric charge).
In practice, the actual level of revenue from tariffs depends on the actual demand
that eventuates. If demand is less than forecast then the business will collect less
revenue; if demand turns out to be higher than expected, the business will collect
more revenue than expected.
The extent to which forecasting errors lead to under- or over-recovery of costs
depends on the extent to which the tariff structure is reflective of the business’
underlying cost structure. In particular, if the volumetric price is set to reflect the
costs that vary with output (i.e. SRMC operating cost), then any change in
demand will be offset by a change in the business’ costs of supply. If the
volumetric price is set at a level that does not reflect the variable costs incurred
by the business, then variations in demand from those forecasts will lead to
revenues diverging from costs, with the extent of this divergence depending on
how far away the volumetric price is from the actual marginal costs of the
business.
Currently, SCA recovers approximately two thirds of its revenue requirement
through its volumetric charge. This is despite the fact SCA’s cost structure is
predominantly fixed – with relatively few costs varying with the level of water
sold to Sydney Water or other customers. In submissions to past determinations,
September 2011 | Frontier Economics 21
Objectives
SCA has pointed out that this introduces considerable revenue volatility when
combined with other factors such as a high degree of uncertainty in demand
forecasting. In 2008, for example, SCA reported that its water sales for the 2005
determination period were 12 per cent less than the forecasts IPART used to set
prices in 2005 due to restrictions. SCA estimated that this resulted in a total
shortfall in revenue of approximately $57 million ($14 million of which relates to
variations greater than the 10 per cent dead band allowed by IPART) (IPART
2009).
IPART (2009) has suggested scarcity pricing (whereby prices vary inversely with
dam levels) could mitigate sales risk.
A pricing approach that varied SCA’s volumetric price inversely with its dam levels
could also mitigate sales risk to SCA. Presently, if SCA’s sales are less than forecast
when setting its volumetric price (e.g. due to the effect of higher than forecast
restriction levels in reducing water demand), it is at risk of under recovering its
revenue requirement – particularly as its costs are mostly fixed. To date, this has
acted to limit the extent to which IPART can increase SCA’s volumetric charge at the
expense of its fixed charge.
Another specific concern identified by SCA in past determinations is that current
volumetric prices do not vary when it incurs additional operating costs when dam
levels fall and it must pump supplies from the Shoalhaven.
Given the high and unpredictable costs associated with pumping, the SCA proposes
that the full cost of future pumping be a pass through component of SCA’s cost
recovery from Sydney Water as it is no longer viable for the SCA to carry these
significant costs, and not equitable for the charges to be borne by all NSW taxpayers
through the resulting reductions in returns (SCA 2009).
Hence, guiding questions in assessing the alternative pricing options include:
● Does the approach provide for sufficient revenue for SCA to finance its
activities?
● Does it ensure against excessive volatility in SCA’s revenues?
● Does the option prevent monopoly rents? How does it deal with over
recovery?
Appropriate allocation of risk
Closely related to the issue of revenue adequacy and stability is the question of
how risk should be allocated between the parties. Of particular relevance here is
how to allocate demand risk between the SCA and Sydney Water.
IPART notes that the allocation of financial risk (between SCA, Sydney Water
and water customers) arising from any new water pricing arrangements is an issue
that requires further consideration.
Currently, SCA has little control over many of the levers that influence demand
risks, such as investing in new supplies to avoid restrictions.
22 Frontier Economics | September 2011
Objectives
Hence, guiding questions in assessing the alternative pricing options include:
● Is risk allocated to the party best able to manage or bear it?
Administrative simplicity
Ease of operation is concerned with ensuring that a pricing approach is practical
to implement. In particular, ease of operation is concerned with ensuring there
are no institutional, administrative or other barriers that would prevent the
approach being implemented. Administrative simplicity means that the resources
required to implement a pricing approach (in terms of administration,
compliance, enforcement and information costs) are proportional to the benefits
of the approach.
There is a range of approaches to implementing scarcity pricing with different
levels of complexity. One key factor will be whether prices relate to increased
operating costs incurred by SCA, which are relatively easy to estimate and
incorporate within the current pricing regime, or based on the opportunity cost
of water in dams, which adds a layer of complexity (e.g. estimating the value of
water and dealing with excess revenues).
Another factor that will influence administrative complexity is the overall design
of the pricing regime (i.e. number of steps in the pricing schedule and frequency
of prices charges).
Guiding questions in assessing this criterion include:
● Is it practical to implement?
● Are there institutional, administrative or other barriers that would prevent an
approach being implemented?
● What are the administrative costs for water service providers, regulators and
customers (and do these costs outweigh the benefits that are likely to accrue
by implementing the approach)?
Transparency
Transparency ensures that water users and others can understand and hence have
confidence in the arrangements. The NWI objectives highlight the importance of
price transparency in water storage and delivery systems (Council of Australian
Governments, 2010).
While scarcity pricing provides a means of providing more cost-reflective and
efficient price signals, there will be a trade-off between the sophistication of the
pricing regime and transparency.
The guiding question in assessing this criterion is whether customers and other
stakeholders readily ascertain and understand what prices are being charged and
how they are determined
September 2011 | Frontier Economics 23
Assessment of options
4 Assessment of options
4.1 Overview
IPART’s 2009 determination for SCA highlighted several options for introducing
scarcity pricing at the wholesale level. These broadly include:
● Scarcity pricing based on SCA’s operating costs – this involves setting
SCA’s variable charge to Sydney Water based on its short-run operating costs
and increasing this charge when dam levels trigger increased operating costs
(particularly due to Shoalhaven pumping cost).
● Scarcity pricing based on the costs of alternative triggered supply and
demand management options – this involves setting a variable charge to
Sydney Water based on estimates the opportunity cost of using dam water
(e.g. the cost of Sydney Water operating the desalination plant or imposing
water restrictions). The price increases would have links to existing operating
rules that require Sydney Water to deploy specific supply or demand
management options when dam levels fall to a certain levels.
● Dynamically efficient pricing based on a system optimisation model
for Sydney – involves using an economic model to calculate a schedule of
efficient prices defined in terms of dam levels.
Below we compare these options to current arrangements.
4.2 Scarcity pricing based on SCA’s operating costs
4.2.1 Description
One option for sending more efficient signals about the marginal cost of using
dam water is to base SCA’s volumetric prices on its short-run operating costs,
which would increase as SCA activates drought supply measures triggered by
dam levels. For example, Shoalhaven drought transfers currently commence
when dam levels fall to 75%. Hence, the volumetric price of dam water could be
say $30/ML (non-drought operating cost) when dam levels were above 75%
storages and $70/ML (marginal pumping costs of Shoalhaven transfer) when
dam levels were below 75%. A similar logic would apply if further decreases in
dam levels triggered even higher marginal operating costs for SCA.
IPART could adjust SCA’s variable price annually (or some other period) to
reflect storage levels at the time. A benefit of this option from SCA’s point of
view is that there would be an inbuilt mechanism to increase the variable price
during the regulatory period, which would mitigate the risk of incurring
additional pumping costs when dam levels fall unexpectedly.
24 Frontier Economics | September 2011
Assessment of options
In its submission to IPART’s 2009 price determination, SCA suggested that
increasing charges to Sydney Water to reflect Shoalhaven pumping costs during
drought was a form of scarcity pricing.
The commencement and cessation of pumping from the Shoalhaven is directly
triggered by dam levels. The SCA contends that passing through SCA's Shoalhaven
pumping cost, if and when it occurs, is consistent with this approach.
A fixed charge would recover any estimated revenue shortfall from SCA’s
variable charge. As SCA’s volumetric prices would relate to its operating costs
(SRMC), which are less than the current volumetric price set with reference to
LRMC, SCA’s fixed charge would account for a larger proportion of SCA’s
revenue than under current arrangements. For example, the fixed fee might
account for around 90% of revenue as opposed to 35%.
The exact level of the fixed charge will depend on SCA required revenue (e.g.
$190m) and expectations about the revenue generated from variable charges,
which will in turn depend on expected water sales and dam levels (e.g. low dam
levels will trigger a higher volumetric price and more revenue per ML of water
sold). If prices are set during a non-drought period when water sales are expected
to be 500 000ML and SCA’s volumetric price is $30/ML, for example, variable
charges will recover $15m and the fixed charge could be set to recover the $175m
shortfall from required revenue.
If predictions are wrong and dam levels and water sales decline during the
regulatory period, the variable charge will increase automatically. IPART could
adjust SCA’s fixed charge when it changes the variable charge or simply set the
fixed charge for the regulatory period and address any under or over recovery at
the next price review. The table below assumes that a fixed charge would apply
for the regulatory period but the variable charge would change as dam levels
change.
Table 1: Illustrative scarcity pricing schedule to Sydney Water (based on SCA
operating costs)
Storage level Scarcity price/Cost Basis
Above 75% $30 per ML
(plus fixed charge, e.g. $175m)
SCA non-drought operating
cost
75% $70 per ML
(plus fixed charge, e.g. $175m)
SCA marginal pumping costs
for Shoalhaven transfers
An alternative option for addressing revenue shortfalls from unexpected
pumping costs during drought is for SCA to simply pass through these costs on
an annual basis, as and when they occur (either through the fixed or variable
charge). However, this is less akin to marginal cost pricing.
September 2011 | Frontier Economics 25
Assessment of options
4.2.2 Assessment
Efficiency and effectiveness
Efficiency of price signal
Economic theory suggests setting SCA’s variable charge with respect to its short-
run marginal costs of operating infrastructure will send an efficient signal to users
of those infrastructure services. As falling dam levels trigger Shoalhaven pumping
costs and a higher volumetric price, for example, Sydney Water would have an
incentive to reduce demand for dam supplies.
Prices based on infrastructure operating costs would not, however, account for
the marginal value of water itself and hence not reflect the full social cost of
using water in SCA storages.
Impacts on Sydney Water’s bulk water sourcing decisions
As the SRMC of operating water supply infrastructure (e.g. ranging from
$30/ML to $70/ML) is lower than the current bulk charge ($250/ML), a move to
more cost-reflective pricing for infrastructure would theoretically provide Sydney
Water with an incentive to use more dam water relative to other supplies when it
was not scarce. In practice, Sydney Water’s existing regulations and desalination
operating rules potentially reduce the extent to which this would happen in
practice (particularly in the short term).
Given the estimated magnitude of the increase in operating costs attributed to
Shoalhaven transfers (e.g. from $30/ML to $70/ML) it is debateable whether a
price increase to reflect increased drought supply costs would loom large in
Sydney Water’s procurement decisions.
Revenue adequacy and stability
The move from setting SCA’s volumetric price with reference to its LRMC to its
SRMC of operating infrastructure would increase SCA’s revenue stability as it
would mean SCA would generate a much higher proportion of revenue from its
fixed charge (e.g. increasing from 35% to 90%) which is not responsive to
changes on water sales. This would be more reflective of SCA’s underlying cost
structure.
Introducing prices that more accurately reflected increased costs associated with
Shoalhaven transfers when dam levels fall would also address revenue adequacy
for SCA. From a cost recovery perspective, drought costs associated with
Shoalhaven transfers would appear to be a potentially significant component of
SCA’s operating costs and hence likely to have a material effect on revenue
adequacy. For example, SCA estimates that unforeseen costs of pumping water
from the Shoalhaven over the 2005 determination period amounted to $31
million to the end of 2007/08, with annual drought pumping costs of $8.5
26 Frontier Economics | September 2011
Assessment of options
million in 2005-06, $9.5 million in 2006-07, and $12.8 million in 2007-08. For
perspective, allowed operating expenditure for the period was $334.3m (or
approximately $81m – $85m annually) (IPART, 2009). This means drought
transfers accounted for approximately 10 per cent of operating costs.
In general, we believe there is a strong case for SCA to be able to recover
efficient costs associated with Shoalhaven transfers when they are required.
Table 2 is an illustrative example of the impact of SRMC pricing on revenue
adequacy for SCA compared to current arrangements under different dam levels
and water sales. It broadly shows how actual revenues from tariffs may diverge
from expected revenue as dam levels and sales fall. In this example, we assume:
● The expected revenue requirement is $190m.
● Water sales vary with dam levels. This might arise because existing operating
rules for Sydney Water reduce its demand for SCA water when dam levels fall
(e.g. desalination operation, restrictions).
● Sales of 500 000ML is the basis for setting the fixed charge to ensure revenue
adequacy and fixed charges are constant for the regulatory period.
● Actual revenues are annual revenues generated from tariffs under different
storage and water sales scenarios (i.e. volumetric price multiplied by sales at a
given dam level plus revenue from the fixed fee). Under scarcity pricing, the
volumetric price will be higher in years when dam levels are below 75% (i.e.
$70/ML instead of $30/ML). For simplicity, we assume a single volumetric
price applies in any one year.
Given these assumptions:
● Under current arrangements, a volumetric charge of $250/ML is expected to
recover $125m when dams are full and annual water sales are 500 00ML (i.e.
500 000ML multiplied by $250/ML) with a fixed charge of $65m recovering
the residual of the revenue requirement (i.e. $190m-$125m)
● Under the SRMC pricing approach, the variable charge (of $30/ML) is
expected to recover $15m (i.e. 500 000ML multiplied by $30/ML) in revenue
when dams are full with a fixed charge of $175m recovering the residual of
the revenue requirement.
Table 2: Indicative revenues from volumetric and fixed charges under different pricing
regimes and dam levels
Annual
Sales
(000’s
Storages
(%) Revenue from charges ($000)
September 2011 | Frontier Economics 27
Assessment of options
ML)
Actual
revenue
Current
volumetric
price
($250/ML)
Current
fixed
charge
Difference
between
actual and
expected
revenue
($190m)
Actual
revenue
SRMC
($30/ML
above 75%,
$70/ML
below 75%)
Higher
fixed
charge
Difference
between
actual and
expected
revenue
($190m)
500 90-100 125 000 65 000 0 15 000 175 000 0
490 80-90 122 500 65 000 -2 500 14 700 175 000 -300
480 75-80 120 000 65 000 -5 000 14 400 175 000 -600
470 70-75 117 500 65 000 -7 500 32 900 175 000 17 900
460 60-70 115 000 65 000 -10 000 32 200 175 000 17 200
450 50-60 112 500 65 000 -12 500 31 500 175 000 16 500
400 Under 50 100 000 65 000 -25 000 28 000 175 000 13 000
Note: This is a simplified example for illustrative purposes only.
As shown in the table, the current pricing approach consistently recovers less
revenue than expected as dam levels and water sales fall, with the annual shortfall
ranging from $2.5m to $25m. In practice, the reduction in water sales will reduce
SCA operating costs, which will partially offset the amount of revenue under-
recovered. For example, the reduction in operating costs when sales are 490
000ML rather than 500 000 ML is $300 000 (i.e. $30/ML multiplied by 10
000ML). However, this leaves a shortfall of $2.2 m.
Under the SRMC pricing approach, the volumetric price initially recovers less
revenue than expected when water sales and dam levels fall (from 100% to 75%
storages). As explained above, however, reduced operating costs due to lower
sales will exactly offset this shortfall between actual and expected revenue from
sales.
Below 75% storage levels, the SRMC pricing approach generates higher revenue
than expected due to the higher volumetric charge (i.e. $70/ML instead of
$30/ML). As all water supplied is priced at the marginal cost of Shoalhaven
transfers, the pricing regime will generally generate revenue sufficient to cover at
least the additional operating costs associated with Shoalhaven transfers.
A potential issue with using dam levels to trigger an increase in the SCA’s
volumetric prices is that the price would apply to all SCA water supplies below
75% storage, regardless of whether it was pumped from the Shoalhaven or not.
This may create a perverse incentive for SCA. For example, SCA could decide to
not pump water when dam levels fall below 75% (and thus not incur additional
operating costs) and it would still earn additional revenue from the increase in the
28 Frontier Economics | September 2011
Assessment of options
volumetric price. In practice, the regulator could claw back this additional
revenue at the end of regulatory period.
An alternative approach to recovering pumping costs would be to include an
annual surcharge to reflect the cost of transfers as and when they occur (i.e. the
price would not be directly linked to dam levels). The cost pass-through could be
averaged across the volumetric price or added as a lump sum to the fixed fee.
This is more aligned with the cost pass-through mechanisms proposed by SCA in
the most recent price determination. A cost pass-through mechanism (based on
average or total pumping costs) may dilute the efficiency properties of marginal
cost pricing, but as noted Sydney Water may have limited flexibility/incentive to
respond to these signals in any event.
Appropriate allocation of risk
A move to SRMC pricing for SCA’s infrastructure services would arguably be
more consistent with efficient risk allocation given SCA currently has few tools
with which to manage sales risks. However, this would increase risks to Sydney
Water. Under current arrangements, reductions in Sydney Water’s revenues from
reduced sales are offset by reductions in purchases from SCA. If SCA charges
Sydney Water a higher fixed charge, however, reductions in Sydney Water’s
revenues from reduced sales would result in a larger shortfall in cost recovery for
Sydney Water.
Administrative complexity
A change to the structure of SCA’s tariffs (balance of fixed and variable charges)
could occur within the existing regime. However, introducing a cost-pass through
mechanism to reflect supply costs may require adjusting prices during the
regulatory period. Administrative costs would increase if Sydney Water was
required to pass on these price signals to customers. Sydney Water has previously
expressed concerns about the administrative costs associated with passing on
Shoalhaven pumping costs to retail customers and suggest prices changes be
limited to once per year.
Sydney Water submitted that it would be concerned if Shoalhaven pumping costs
were to be passed through to its customers immediately after they were incurred (i.e.,
at the next bill). It opposed this approach as it could involve up to four price changes
to its customers in a year, and result in high administrative costs to reconfigure billing
systems and inform customers. However, Sydney Water indicated that if IPART
assesses that SCA is not able to absorb Shoalhaven pumping costs in between
determinations, the pass-through mechanism in Sydney Water’s determination could
be used to pass through these costs on an annual basis (IPART 2009).
Transparency
As SRMC pricing on based on operating costs would relate to costs incurred by
SCA, it is arguably more transparent than the current approach of setting the
September 2011 | Frontier Economics 29
Assessment of options
volumetric price somewhat arbitrarily between SRMC and LRMC based on a mix
of different objectives.
Overall
While pricing based on short-run operating costs is consistent with efficient
pricing, using dam levels as the trigger for price increases in SCA’s volumetric
price may create perverse incentives for SCA as it would be able to earn
additional revenue regardless of whether it actually incurred pumping costs. In
practice, the regulator could claw back this additional revenue at the end of
regulatory period.
An alternative would be to include a surcharge to volumetric prices to reflect the
average cost of transfers as and then when they occur. While it is debateable
whether this would affect Sydney Water’s procurement decisions it would help
SCA recover its efficient expenditure in providing drought supplies.
By itself, this option would not reflect the full cost of Sydney Water drawing on
dam water as it excludes the marginal value of water itself.
4.3 Scarcity pricing based on the cost of alternative
triggered supply and demand options
4.3.1 Description
Another option for using prices to reflect the social marginal cost of using dam
water is to set prices that reflect the marginal value of water at different dam
levels based on the cost of alternative options triggered. For example, under
current operating rules the Sydney Water desalination plant begins operation
when dam levels fall to 70%. This would suggest that the marginal value of water
is at least equal to the marginal cost of producing desalination water. To send
signals to Sydney Water about the cost of using dam water, SCA’s volumetric
price for water could potentially increase by an amount commensurate with the
operating cost of desalination (adjusted for avoided system costs etc).
Similarly, SCA or IPART could assign a value to the social marginal cost of water
restrictions when dam level fall to a certain trigger levels (e.g. 50%). IPART
(2009) appears to allude to this option when it notes ‘under a scarcity pricing
approach, higher level water restrictions (as a result of low dam levels) will result
in proportionally higher volumetric SCA prices’.
Table 3 shows an illustrative schedule of scarcity prices and triggers that
SCA/IPART could potentially apply. The pricing schedule could have more or
less steps depending on the availability of options and cost estimates. The price
assigned to the cost of water restrictions is a hypothetical cost for illustrative
purposes.
30 Frontier Economics | September 2011
Assessment of options
These prices would be in addition to, but separate from, SCA’s charges to
recover infrastructure costs discussed in section 4.2.1. As discussed below, the
revenue from scarcity charges based on the opportunity cost of water would not
necessarily accrue to SCA.
Table 3: Illustrative scarcity pricing schedule to Sydney Water - cost of alternative
triggered supply and demand options
Storage level Scarcity price/Cost Basis
Above 70% Marginal value of water $0/ML
(plus SCA infrastructure charges)
Full storages imply low
scarcity value of water
70% $422 per ML
(plus SCA infrastructure charges)
Marginal cost of operating
desalination
50% $650 per ML (hypothetical)
(plus SCA infrastructure charges)
Marginal cost of prolonging
water restrictions
SCA or IPART could potentially develop a smoothed pricing schedule based on
these estimates (table 4). This approach may require defining steps based on the
average cost of two options (i.e. the marginal cost of the present option and the
marginal cost next option) or assigning probability weights as falling dam levels
increase the chance of incurring costs associated with triggering the next source.
This would avoid large price changes as a trigger point is reached.
Table 4: Illustrative scarcity pricing schedule to Sydney Water (smoothed prices)
Storage level Scarcity price/Cost Basis
90% Marginal value of water $0/ML
(plus SCA infrastructure charges)
Full storages imply low scarcity value
of water
80% $211 per ML
(plus SCA infrastructure charges) Half the cost of operating desalination
70% $422 per ML
(plus SCA infrastructure charges) Cost of operating desalination
60% $536 per ML
(plus SCA infrastructure charges)
Mid-point of the cost of operating
desalination and cost of prolonging
water restrictions
50% $650 per ML (hypothetical)
(plus SCA infrastructure charges) Cost of prolonging water restrictions
An alternative approach is to estimate a schedule of scarcity prices using
sophisticated modelling techniques, such as stochastic dynamic programming,
September 2011 | Frontier Economics 31
Assessment of options
that considers pricing, storages, and investment decisions in an integrated way
(this is discussed as a separate option on 4.4 below).
4.3.2 Assessment
Efficiency and effectiveness
Efficiency of price signal
Defining robust scarcity prices that reflect the marginal value of water in dams
can present significant theoretical and practical challenges. While a price based on
the cost of desalination could act as a proxy for the opportunity cost of using
dam water, for example, the regulator would need to consider whether a
relatively basic regime (as described above) would be sufficient to send an
efficient signal for investment or whether a more sophisticated pricing approach
would be required. The Independent Panels for the Metropolitan Water Plan
suggested that that a ‘shadow price’ based on desalination operating costs would
not fully capture the system wide opportunity costs associated with substituting
desalinated water for dam water.
The harder the plant is operated on average, the less air space is available in
storages and there is a reduction in the value of this harvesting option. This is a cost
that needs to be added to obtain the social operating costs of operating the plant.
Further information would be required to calculate an adjusted shadow price along
these lines.
Another issue when estimating the opportunity cost of water based on the
operating costs of desalination (or other technologies) is determining what the
scarcity price should be when production reaches full capacity but dam levels
continue to decline. That is, the opportunity cost may exceed the operating cost
of desalination following a sequence of low inflow months or years. Notably, the
desalination plant produces up to 90GL per year while dam levels can fall up to
five hundred GL per year (see Figure 4).
Estimating the cost of water use in prolonging restrictions (triggered when dam
levels are below 50%) could draw on studies on the costs of water restrictions
(e.g. willingness to pay studies) incorporated in the recent review of desalination
plant operating rules (CIE 2010). However, including these costs in a simple
pricing schedule based on dam levels would require converting them into costs
per megalitre of use.
Given the operating rules for the desalination plant take into account the risk of
restrictions, it is arguable that Sydney Water already internalises these costs to
some degree.
Impacts on Sydney Water’s bulk water sourcing decisions
32 Frontier Economics | September 2011
Assessment of options
The magnitude of price increases related to scarcity pricing is potentially
significant (e.g. $422/ML for desalination) and is therefore much more likely to
influence Sydney Water’s water sourcing and investment decisions than cost-
reflective pricing for raw water operating costs alone.
Revenue adequacy and stability
Scarcity prices based on the opportunity cost of water (e.g. desalination costs,
water restrictions) would generate additional revenues without SCA incurring a
corresponding cost (i.e. it is a resource rent). These revenues could potentially be
substantial. For example, annual sales to Sydney of 450 000 ML per year (IPART,
2009) and a scarcity price of $422/ML would generate approximately $190m in
revenue per year (assuming dam levels were below the trigger level the whole
time).
The table below presents information on revenues from pricing SCA’s
infrastructure services at SRMC from section 4.2.2 along with annual revenues
from the two illustrative scarcity pricing models. Notably the revenues generated
by scarcity charges exceed SCA’s fixed infrastructure charge to Sydney Water
when dam levels fall below 70%.
Table 5: Indicative revenues from volumetric charges under different pricing regimes
and dam levels
Annual
Sales
(000’s ML)
Storages
(%)
Revenue from charges ($000)
SRMC
($30/ML
above 75%,
$70/ML
below 75%
Fixed
Difference
from expected
revenue
($190m)
Scarcity
pricing
Scarcity
pricing
(smoothed)
500 90-100 15 000 175 000 0 0 0
490 80-90 14 700 175 000 -300 0 0
480 75-80 14 400 175 000 -600 0 101 280
470 70-75 32 900 175 000 17 900 0 99 170
460 60-70 32 200 175 000 17 200 194 120 194 120
450 50-60 31 500 175 000 16 500 189 900 241 200
400 Under 50 28 000 175 000 13 000 260 000 260 000
Note: This is a simplified example for illustrative purposes only.
This raises the key issue of what happens to the revenues from the scarcity
charge. One option would be for SCA to return this revenue to government (i.e.
September 2011 | Frontier Economics 33
Assessment of options
as a resource rent tax). With respect to the potential for over-recovery of scarce
resources, a recent Productivity Commission staff working paper (Barker,
Murray, & Salerian, 2010) advocated a resource rent tax rather than regulating
returns:
Capacity rents are distinct from monopoly rents and have different implications.
Monopoly rents arise from exploiting market power, creating costs to community
Capacity rents, on the other hand, accrue to the owners of capacity-constrained
resources (such as aquifers), and act to ration limited supply so as to achieve an
efficient market equilibrium. Whereas the existence of monopoly rents might mean
there is a role for government regulation to address market power, capacity rents
should not be regulated away. Where firms make excessive profits as a
consequence of capacity rents, this can be addressed more efficiently through
resource-rent taxation that does not distort the price of water.
Alternatively, IPART could use some of the revenue to decrease the fixed charge
to Sydney Water on the basis this would result in lower costs to customers/the
public and offset the increase in overall variable charges to Sydney Water
(including SCA infrastructure costs and the opportunity cost of water). Given
there would potentially be a link between Sydney Water’s current operating
decisions, which influence supply scarcity, and the size of the fixed charge rebate
in the next period received by Sydney Water, this approach may have unintended
consequences on Sydney Water’s sourcing and investment decisions.
IPART could also adjust SCA’s required revenues in the next regulatory period
to account for excess revenues reflecting scarcity rent.
To the extent SCA’s infrastructure prices are set according to its short-run
operating costs (see above) with an appropriate fixed charge, its revenues should
be sufficient to cover costs and provide revenue stability without the additional
scarcity rents from water.12 In fact, if SCA did collect revenue from the scarcity
charge its revenues could potentially be more volatile and more difficult to
forecast in a regulatory setting (i.e. variable prices would be $422/ML rather than
$250/ML). Separating resource pricing and infrastructure pricing mitigates this
effect.
Appropriate allocation of risk
Compared to current arrangements, scarcity pricing would increase bulk supply
costs to Sydney Water when water became scarce. This would present financial
risks to Sydney water given current retail pricing arrangements are invariant to
dam levels and dams remain its main source of supply. In the longer term,
however, Sydney Water has tools to manage these risks such as investing in new
supplies and demand management and changing the retail price structure or
12 As noted earlier, leaving fixed charges as they currently stand would not achieve revenue adequacy.
34 Frontier Economics | September 2011
Assessment of options
levels. As noted, IPART could use some of the revenue from the scarcity charge
to offset Sydney Water’s fixed charge.
There may be a case for introducing additional steps to smooth prices. This
would avoid large price spikes to Sydney Water (e.g. from $70/ML to $422/ML)
while still providing an incentive to respond to prices over the short to medium
term. It would also potentially help signal the increased likelihood of incurring
higher costs.
Administrative complexity
Introducing a simple scarcity charge with a small number of prices/triggers and
that only changed infrequently (e.g. less than once a year) would arguably be
administratively feasible, particularly given it essentially only applies to one large
customer (although it will be important to address specific issues for the three
smaller local councils). This pricing option has some similarities to the ACT
water abstraction charge, whereby water provider ACTEW faces a volumetric
charge for water use (0.55c/KL) that, among other things, is claimed to reflect
the scarcity value of water. The ACT government collects the revenue from this
charge (ACTEW, 2011).
A more sophisticated regime, with multiple scarcity prices/triggers, which
required frequent price changes, would add to the administrative burden.
A major issue will be developing an appropriate proxy for the value of water in
storage that stakeholder can agree upon. Notably, there have been (unsuccessful)
legal challenges of the ACT water abstraction charge.
Transparency
In contrast to water markets, administered scarcity pricing requires estimating
values associated with water, which rely on a number of assumptions. By its
nature, this process can be highly contentious. However, links to the costs of
alternatives at least provides a defensible benchmark.
Under current arrangements, the regulator is already implicitly estimating these
values when setting the volumetric charge. The basis for calculating the LRMC is
often not transparent.
Overall
Setting a charge based on the marginal value of water in dams (which is separable
from a charge to recover SCA’s short-run operating cost) could reduce the
revenue risk to SCA embodied in current tariff structures while still providing a
more efficient price signal for sourcing decisions by Sydney Water.
A relatively basic approach to valuing water in storage, such as setting prices
based on the cost of desalination (when operating rules based on dam levels
trigger operation of the desalination plant), may achieve the broad aim of sending
September 2011 | Frontier Economics 35
Assessment of options
a price signal to Sydney Water. SCA could introduce additional steps in the
pricing regime to smooth prices to reduce price volatility to Sydney Water.
Developing an appropriate proxy for the value of water in storage, which
stakeholders agree upon, is likely to be a key challenge.
4.4 Scarcity pricing based on a system optimisation
model
4.4.1 Description
Instead of the heuristic approaches above, IPART could adopt a more
sophisticated approach. In particular, it could use economic modelling techniques
such as stochastic dynamic programming to define a schedule of prices that
optimise community welfare given expected policy constraints, planned
investments, future inflows, storages and demand forecasts.
In theory, it should be possible to define a schedule of efficient prices in terms of
dam levels, which could be updated either annually or at the start of the
regulatory period to account for new information on rainfall, inflows and dam
levels. The regulator (drawing on expert advice) would need to make judgments
about whether the assumptions underpinning modelled scarcity prices (e.g.
chosen inflow scenarios) are reasonable.
Figure 10 illustrates modelled scarcity-based prices under a range of
rainfall/inflow scenarios. As these prices relate to end-consumer prices provided
by a vertically integrated service provider, further consideration would need to be
given to applying the same principles to SCA wholesale operations and how this
fits within the broader Sydney supply system. For example, Sydney Water’s
decisions on how to operate the desalination plant would influence the optimal
pricing, storage, and investment patterns of SCA as it would affect demand. To
the extent SCA and Sydney Water are subject to operating rules, however, these
might be included as constraints in the model.
In theory, however, an increase in SCA’s wholesale prices should have an effect
on the optimal operating rules for the desalination plant. That is, the review of
desalination plant operating rules by CIE estimated costs and benefits of
alternative operating regimes assuming scarcity pricing had no role in managing
supplies. If SCA introduced scarcity pricing, this may affect the estimated costs
and benefits of alternative operating rules and their relative rankings.
36 Frontier Economics | September 2011
Assessment of options
Figure 10: Prices under scarcity-based pricing, across various rainfall scenarios
Source: Barker, Murray, & Salerian, 2010.
In Western Australia, the ERA (2009) has developed a model to estimate the
short-run value of water. The model is based on a hypothetical wholesale market
for metropolitan water supply. The model calculates the price at which supply
equals demand for each of the next five years, given the available supply options,
supply security requirements and an assumption about the responsiveness of
demand to price (ERA 2009). The ERA notes that the model is useful as one
source of information for price setting purposes.
The ERA model does not seek to attach a value to water in storage through
specific scarcity prices. Instead, it defines a demand schedule for bulk water that
achieves ‘the amount of water that would ideally be retained in the dams at the
end of each year to secure the system’. For example, the security target is to
retain enough water in the dams at the end of each year to ensure ‘saturated’
demand13 will be met in the following year even if zero inflows occur.
4.4.2 Assessment
Efficiency and effectiveness
Efficiency of price signals
Although economic modelling to generate a dynamically efficient scarcity price
represents a theoretically attractive approach for efficient pricing, its practical
13 Saturated demand is defined as 30 per cent above the level of demand that would occur under a
total sprinkler ban.
September 2011 | Frontier Economics 37
Assessment of options
application in the water sector is in its early stages. Further, such models need to
be tailored to specific systems or policy questions.
By their nature economics models rely on value judgements and technical
assumptions and are information intensive. The quality of the model design,
inputs and application will therefore be a key factor influencing whether
modelled prices are likely to enhance efficiency. That said, existing pricing
arrangements also embody similar types of value judgements and technical
assumptions and require similar information.
Overall, however, modelled scarcity prices would be expected to lead to more
efficient pricing than the status quo. Modelling by Barker, Murray, & Salerian,
(2010) for example showed that scarcity pricing was associated with higher
community welfare than pricing based on LRMC.
Impacts on Sydney Water sourcing decisions
Although prices modelled by the Productivity Commission are broadly indicative
at best, they suggest prices could diverge significantly from the short-run
marginal cost of supplying and distributing water from dams (i.e. minimum price)
depending on rainfall and inflows. Largely, the estimated price rise remains
within a relatively narrow band 90 per cent of the time. Under more extreme
scenarios, the modelled scarcity price rises to many times the short-run marginal
(operating) costs. Such prices may well be material to Sydney Water’s
procurement decisions under drier scenarios.
As noted above, the introduction of scarcity pricing may change the optimal
operating rules for the desalination plant and hence Sydney Water’s sourcing
decisions.
Revenue adequacy and stability
Recent theoretical applications of economic modelling to estimate scarcity prices
tend to abstract from institutional issues such as price regulation and tariff
structures or deal with them in a general way. The Productivity Commission
(Barker, Murray, & Salerian, 2010), for example, notes ‘fixed charges (under a
two part tariff) are not included in the modelling undertaken for this study’. To
the extent this model is an extension of the SRMC scarcity pricing models
described above, it should enable SCA to achieve revenue adequacy.
Appropriate allocation of risk
As with other forms of scarcity pricing, a common concern about economic
modelling approaches to efficient pricing is the fluctuation in prices. In response,
the ERA notes that price variations implied by its wholesale level model would
not necessarily translate to similar changes in retail prices and could actually help
identify required prices over a set regulatory period.
38 Frontier Economics | September 2011
Assessment of options
The submission from the Water Corporation that a short run water model produces
prices that fluctuate from year to year does not mean that usage charges need to
fluctuate to the same extent from year to year. Indeed, the model is likely to be most
useful if it is used to identify the value of water (and hence usage charges) over the
course of the regulatory period (ERA 2009).
Administrative complexity
As an initial step, the NSW government, SCA or IPART could develop a model
specific to the Sydney system (likely drawing on existing models, such as
WATHNET and economic models developed to assess options under the
metropolitan plan) that considers scarcity pricing in a system wide context. This
would help establish the workability of this approach to setting scarcity prices
and identify any interactions between optimal scarcity prices and other operating
rules.
Transparency
One key issue associated with economic modelling is whether stakeholders, such
as the regulator, Sydney Water and its customers, would be willing to accept
prices determined though this complex ‘black box’ process. In Western Australia,
for example, there has been much debate about the specification of the ERA’s
proposed short-run value of water model:
From a practical perspective, the ERA’s proposed SRMCP model is not well
specified, calibrated or tested, and provides highly unstable results under a wide
range of foreseeable circumstances. Without a strong theoretical driver, adopting a
methodology that has a high probability of being abandoned at the next price review
(due to the potential for unreasonably high or low prices) is not good regulatory
practice (Water Corporation submission on Draft Report, Part A, cited in ERA 2009)
There is a risk that such modelling would be simply set aside or rejected when it
provides results that stakeholders do not agree with and which IPART cannot
easily communicate.
Overall
In theory, an economic model that calculated an optimal scarcity price based on
existing system constraints, planned investments and operating plans would
produce efficient price signals. However, much work would be required to
develop such a model and to achieve buy-in from stakeholders. Further,
consideration needs to be given to how these models would interact with the
models underpinning the existing operating rules for desalination.
September 2011 | Frontier Economics 39
Conclusions
5 Conclusions
In considering options for scarcity pricing at the wholesale level, stakeholders will
approach the issue with different perspectives. As a business, a primary objective
of SCA is to ensure its prices to Sydney Water enable it to earn sufficient revenue
to fulfil its supply functions, including managing its infrastructure efficiently and
in accordance with sound commercial principles. IPART, on the other hand,
must consider broader issues relating to the efficient operation of the supply
system, consisting of multiple sources, and the possible development of a
wholesale market. In particular, it must consider whether SCA’s prices provide
sufficient incentives for Sydney Water to undertake efficient investment in,
alternative supplies and demand management.
Under current pricing arrangements, there is an inherent conflict between
achieving revenue stability for SCA on the one hand and IPART’s desire to send
a signal on the value of water to inform Sydney Water’s sourcing and investment
decisions on the other. A key issue in assessing options for scarcity pricing is
whether they can provide appropriate signals for efficient use and investment
while not exposing SCA to undue revenue risk. There is a strong argument that
SCA, as a relatively passive manager of catchments and dams in accordance with
government defined operating strategies, should be able to recover its efficient
costs without being excessively exposed to demand risk over which it has no
control.
Our view is that scarcity pricing (based on principles of marginal cost pricing)
would address this issue. This model would broadly involve:
● Aligning SCA’s volumetric price with its short-run operating costs, with a
fixed charge to address any revenue shortfall. This would require a large
increase in revenue generated from SCA’s fixed charge compared to current
arrangements.
● Setting a separate volumetric price that reflects the estimated marginal value
of water in storage, which would be in addition to SCA’s infrastructure
charges but would effectively apply only when predefined triggers are
reached.
The revenue collected through this additional scarcity charge represents a
separate resource rent outside of SCA’s required revenue requirement
and could potentially be retained by government or alternatively used to
offset fixed Sydney Water’s charges in the current or future regulatory
periods.
This pricing approach would reduce the revenue risk to SCA embodied in
current tariff structures while providing an appropriate price signal for
consumption and investment to Sydney Water (and other SCA customers) and
potential new entrants at times when water is scarce. It does so by clearly
40 Frontier Economics | September 2011
Conclusions
differentiating between pricing for SCA’s infrastructure services and the water
resource itself.
The extent to which these prices will have a material effect on the current and
future portfolio of supply options for Sydney will partly depend on broader
institutional arrangements for urban water planning. For example, current
government policies (for example, strict desalination plant operating rules) may
lock in particular sourcing decisions and investments and thus limit Sydney
Water’s flexibility to respond to wholesale water prices.
Another advantage of this approach is that it would be more flexible to adjust to
changing circumstances in the urban water industry under which Sydney Water
must make increasingly complex water sourcing and investment decisions across
a range of supply and demand management options.
The design of the scarcity pricing regime also needs to consider carefully how
scarcity prices will combine with existing operating rules for the Sydney system to
achieve the most efficient mix of options for balancing supply and demand. For
example, to the extent the operating rules for the desalination plant already take
into account risks to existing storages and therefore reflect an ‘optimal’ operating
strategy, there is a risk that the addition of scarcity pricing for SCA supplies will
double count the costs of consuming dam water. This is not so say scarcity
pricing is not worthwhile, but rather that current operating rules may need to be
reconsidered in light of this new option for balancing supply and demand.
Putting this pricing model into practice will be challenging. In particular,
estimating the value of water in storage is a key issue to determine. In this paper
we canvass potential options for estimating the marginal value of water in storage
ranging from heuristic approaches based on existing operating rules (e.g. setting
prices equal to the operating cost of alternative options such as desalination when
dam levels trigger operation of the desalination plant) to economic modelling
approaches.
In theory, an economic model that calculated an optimal price based on existing
system constraints, planned investments and operating plans would produce
efficient price signals. However, we recognise that much work would be required
to develop such a model and for it to be accepted in a regulatory price setting
context.
In summary, we would advocate replacing the current pricing arrangements
based on setting SCA’s volumetric charges with regard to its LRMC with a more
cost-reflective approach based on SCA’s SRMC together with an additional
scarcity price based on the costs of predefined triggered alternatives. This would
better protect SCA’s revenue adequacy while also achieving IPART’s aim of a
more efficient price signal to SCA’s customers and potential new suppliers.
September 2011 | Frontier Economics 41
<References
References
ACTEW. (2011). Water Abstraction Charge. Retrieved February 2, 2011, from
ACTEW: http://www.actewagl.com.au/water/networks/wac.aspx
Barker, A., Murray, T., & Salerian, J. (2010). Developing a Partial Equilibrium Model o
fan Urban Water System. Melbourne: Productivity Commission Staff Working
Pape.
CIE . (2010). Review of operating regime for Sydney Water’s desalination plant. Sydney:
Centre of International Economics.
Council of Australian Governments. (2004). Intergovernmental Agreement on a
National Water Initiative.
Council of Australian Governments. (2010). NWI Pricing Principles.
ERA. (2009). Inquiry into Tariffs of the Water Corporation, Aqwest and Busselton Water.
Perth: Economic Regulation Authority of Western Australia.
Griffin, R. (2006). Water resources economics: Analysis of scarcity, policy and projects.
London: MIT Press.
Independent Advisory Panel. (2008). Submission from the Independent Advisory Panel
for the Sydney Metro Water Plan, IPART inquiry into the SCA price path, October 2008.
IPART. (2004). Investigation into Price Structures to Reduce the Demand for Water in the
Sydney Basin. Sydney: Independent Pricing and Regualtory Tribunal.
IPART. (2009). Review of prices for the Sydney Catchment Authority From 1 July 2009 to
30 June 2012, Water — Determination and Final Report (June 2009). Sydney: IPART.
Melbourne Water. (2010). Melbourne Water Storages. Retrieved February 2, 2010,
from Melbourne Water :
http://www.melbournewater.com.au/content/water_storages/water_report/wat
er_report.asp?bhcp=1
NSW Office of Water. (2010). 2010 Metropolitan Water Plan. Sydney: Department
of Environment, Climate Change and Water.
Productivity Commission. (2008). Towards urban water reform: A discussion paper.
Melbourne: Productivity Commission.
PwC. (2010). Review of Urban Water Security Strategies. Prepared by
PricewaterhouseCoopers (PwC) at the request of Infrastructure Australia.
SCA. (2011). Operating licence. Retrieved February 2, 2011, from Sydney
Catchment Authority: http://www.sca.nsw.gov.au/publications/publications/5
SCA. (2009). Response to IPART’s Review of prices for the Sydney Catchment Authority:
Draft Determination and Draft Report. Sydney: Sydney Catchment Authority.
42 Frontier Economics | September 2011
<References
SCA. (2011). Water storage and supply reports. Retrieved February 2, 2011, from
Sydney Water: http://www.sca.nsw.gov.au/dams-and-water/weekly-storage-and-
supply-reports
Sydney Water. (2011). Overall project documentation: Desalination. Retrieved February
2, 2011, from Sydney Water:
http://www.sydneywater.com.au/Water4Life/desalination/overalldocumentatio
n.cfm
Sydney Water. (2010). Water Conservation and Recycling Implementation Report 2009-10.
Sydney: Syndey Water.
Frontier Economics Pty Ltd in Australia is a member of the Frontier Economics network, which
consists of separate companies based in Australia (Brisbane, Melbourne & Sydney) and Europe
(Brussels, Cologne, London & Madrid). The companies are independently owned, and legal
commitments entered into by any one company do not impose any obligations on other companies
in the network. All views expressed in this document are the views of Frontier Economics Pty Ltd.
Disclaimer
None of Frontier Economics Pty Ltd (including the directors and employees) make any representation or warranty as to the accuracy or completeness of this report. Nor shall they have any liability (whether arising from negligence or otherwise) for any representations (express or implied) or information contained in, or for any omissions from, the report or any written or oral communications transmitted in the course of the project.
FRONTIER ECONOMICS NETWORK
BRISBANE | MELBOURNE | SYDNEY | BRUSSELS | COLOGNE | LONDON | MADRID
Frontier Economics Pty Ltd 395 Collins Street Melbourne Victoria 3000
Tel: +61 (0)3 9620 4488 Fax: +61 (0)3 9620 4499 www.frontier-economics.com
ACN: 087 553 124 ABN: 13 087 553 124